AgentSkillsCN

github-actions-debugging

指导开发者调试失败的GitHub Actions工作流。当您需要调试GitHub Actions工作流时，可使用此技能。

SKILL.md
--- frontmatter
name: github-actions-debugging
description: Guide for debugging failing GitHub Actions workflows. Use this when asked to debug failing GitHub Actions workflows.
license: MIT

GitHub Actions Debugging Skill

This skill provides a systematic approach to debugging failing GitHub Actions workflows in pull requests.

When to Use This Skill

  • •When a GitHub Actions workflow fails in a pull request
  • •When investigating CI/CD pipeline failures
  • •When asked to analyze workflow run errors
  • •When troubleshooting build, test, or deployment issues

Debugging Process

1. Identify Failed Workflows

Use the list_workflow_runs tool to look up recent workflow runs for the pull request and their status:

code
list_workflow_runs(owner="owner", repo="repo", resource_id="workflow_file.yml")

This will show:

  • •Workflow run status (success, failure, cancelled)
  • •Trigger event and branch
  • •Run timestamps
  • •Run IDs for further investigation

2. Get AI-Powered Failure Summary

Use the summarize_job_log_failures tool to get an AI summary of the logs for failed jobs:

code
summarize_job_log_failures(owner="owner", repo="repo", run_id=12345)

This provides:

  • •Concise summary of what went wrong
  • •Key error messages
  • •Affected jobs and steps
  • •No context window bloat from thousands of log lines

3. Deep Dive Into Logs (If Needed)

If you need more detailed information, use these tools:

For specific job logs:

code
get_job_logs(owner="owner", repo="repo", job_id=67890)

For complete workflow run logs:

code
get_workflow_run_logs(owner="owner", repo="repo", run_id=12345)

4. Reproduce Locally

Try to reproduce the failure in your own environment:

  • •Check out the failing branch
  • •Run the same commands locally
  • •Compare environment differences
  • •Look for platform-specific issues

5. Fix and Verify

Once you understand the issue:

  1. •Make the necessary code changes
  2. •If you reproduced locally, verify the fix works
  3. •Commit your changes
  4. •Monitor the workflow run to ensure it passes

Common Failure Patterns

Build Failures

  • •Missing dependencies
  • •Compilation errors
  • •Configuration issues
  • •Environment variable problems

Test Failures

  • •Flaky tests
  • •Test environment setup issues
  • •Race conditions
  • •Missing test data

Deployment Failures

  • •Authentication/permission issues
  • •Resource unavailability
  • •Configuration mismatches
  • •Network connectivity problems

Best Practices

  1. •Start with the summary - Don't load full logs until necessary
  2. •Check recent changes - Compare with previous successful runs
  3. •Look for patterns - Similar failures across multiple jobs
  4. •Check workflow file - YAML syntax and configuration errors
  5. •Verify secrets - Ensure all required secrets are configured
  6. •Test locally first - Validate fixes before pushing

Tools Reference

  • •list_workflow_runs - List recent workflow runs and their status
  • •summarize_job_log_failures - Get AI summary of failed job logs
  • •get_job_logs - Get detailed logs for a specific job
  • •get_workflow_run_logs - Get complete workflow run logs
  • •list_workflow_jobs - List all jobs in a workflow run
  • •get_workflow_run_usage - Get timing and billable minutes

Example Workflow

code
1. User reports: "The CI is failing on PR #123"
2. List workflows: list_workflow_runs to find failed runs
3. Summarize failures: summarize_job_log_failures for quick diagnosis
4. If needed, get detailed logs: get_job_logs for specific failures
5. Reproduce issue locally and test fix
6. Commit fix with clear message explaining the issue
7. Monitor new workflow run to confirm fix

Remember: Always start with high-level summaries before diving into detailed logs to keep your context efficient.